1. Field of the Invention
The invention relates to optical waveform observing apparatus for observing the waveform of an incident light beam.
2. Description of the Prior Art
One example of conventional means for observing the waveform of an optical pulse having a constant repetitive frequency is an optical waveform observing apparatus of the type including a streak tube used to convert the variation with time of an optical pulse into a luminance distribution, such as a streak image on a fluorescent surface.
FIG. 3 is a block diagram showing the arrangement of a conventional optical waveform observing apparatus as disclosed in U.S. Pat. No. 4,645,918. In that apparatus, a hematoporphyrin derivative 104 is excited with a pulsed light beam repetitively outputted by a dye laser oscillator 101 and emits fluorescence having a waveform that is observed with a streak tube 130. Streak tube 130, as shown in FIG. 3, comprises a photocathode 131 for receiving the fluorescence, an accelerating electrode 135 for accelerating an electron beam outputted by the photocathode 131, a focusing electrode 136 for focusing the electron beam accelerated by the accelerating electrode 135, an aperture electrode 137, deflecting electrodes 133 for deflecting in a sweep mode the electron beam that has been focused by focusing electrode 136 and passed through aperture electrode 137, a microchannel plate 132 for multiplying the deflected electron beam, and a phosphor screen 14, to which the output electron beam of the microchannel plate 132 is applied.
The pulsed light beam repetitively outputted by the dye laser oscillator 101 is divided into two pulsed light beams by a beam splitter 102. One of the two pulsed light beams excites the hematoporphyrin derivative 104 repetitively. The fluorescence emitted by derivative 104 is applied through an optical system 116 to the photocathode surface 131 of the streak tube 130 at a certain repetitive frequency. The other pulsed light beam is applied to a photodiode 105 to form an electrical signal TR for a predetermined deflecting voltage that is applied to the deflecting electrodes 133 in streak tube 130. More specifically, the pulsed light beam is subjected to photo-electric conversion by the photodiode 105 to form the electrical signal TR. The electrical signal TR, being delayed by a time sweep circuit 107 under the control of a control unit 110, is converted into a deflecting trigger signal that is applied to a deflecting circuit 108. Circuit 108 in turn produces a deflecting voltage in synchronism with the deflecting trigger signal.
With respect to the streak tube 130, the output electron beam of the photocathode 131 is deflected by the deflecting voltage applied across the deflecting electrodes 133 in the sweep mode, so that the variation with time of the fluorescence applied to the photocathode 131 is converted into a spatial luminance distribution on The phosphor screen 134, i.e., it is observed as a streak image. In the optical waveform observing apparatus of FIG. 3, the light of the luminance distribution on the phosphor screen 134, i.e., the light of the streak image, is applied through a lens 118 to a sampling board 111 with a slit 109, so that it is sampled by slit 109 to provide a sampling waveform. The sampling waveform is applied to a photomultiplier 112 that subjects it to photo-electric conversion and multiplication. The output of the photomultiplier 112 is applied through an amplifier 113 to a display unit 114.
The operation of the optical waveform observing apparatus thus organized will be described with reference to FIG. 4 which is a time chart consisting of parts (a) through (e).
In response to the pulsed light beam repetitively outputted by the dye laser oscillator 101, the hematoporphyrin derivative 104 outputs fluorescence with a repetitive period. This is indicated in part (a) of FIG. 4 which shows the waveform of an incident light beam "IN" that is applied to the photocathode 131 of the streak tube 130. Also in response to the pulsed light beam of dye laser oscillator 101, the photodiode 105 outputs the electrical trigger signal TR as shown in part (b) of FIG. 4. As is apparent from parts (a) and (b) of FIG. 4, the electrical trigger signal TR is completely synchronized with the incident light beam IN. The electrical trigger signal TR is gradually delayed by time sweep circuit 107 as shown in part (c) of FIG. 4, and converted into the deflecting trigger signal. Under the control of the control circuit 110, the time sweep circuit 107 forms a sampling sequence type deflecting trigger signal that allows a sampling operation at a sampling time, the next sampling operation at the next sampling time, and so on. As shown in part (c) of FIG. 4, at the n-th sampling time, the deflecting trigger signal TR; at the (n+1)-th sampling time, the deflecting trigger signal TR2 is delayed by a period of time (n+1).multidot.t; at the (n+2)-th sampling time, the deflecting trigger signal TR3 is delayed by a period of time (n+2)t; and so forth, where t is the unitary delay time of the deflecting trigger signal.
When the deflecting trigger signals are applied to deflecting circuit 108, the latter produces the deflecting voltage V as shown in part (d) of FIG. 4. As is apparent from part (d) of FIG. 4, the deflecting voltage V is maintained at a potential V.sub.m when it is not used to deflect the output electron beam of the photocathode 131. However, when the deflecting trigger signal is applied to deflecting circuit 108, the deflecting voltage drops to a potential -V.sub.m substantially with a ramp characteristic, to deflect the electron beam. Whenever the deflecting voltage is decreased in the above-described manner, the electron beam is deflected downwardly, as a result of which it is observed as streak images FG1, FG2, FG3, . . . at the sampling times as shown in part (e) of FIG. 4. As a result, the variation in intensity of the incident light beam IN can be observed as a luminance distribution. As seen from parts (a) and (e) of FIG. 4, the streak images FG1, FG2, FG3, . . . reflect the respective incident light beams IN. The streak images are shifted in phase from one another, however, because the deflecting voltages at the sampling times are shifted in phase from one another. When, during the electron beam deflecting operation, the deflecting voltage is near zero (0) volts, the electron beam 20 not deflected and reaches the central portion of the phosphor screen 134, which portion emits a light beam. Most of the light beam thus emitted passes through the slit 109 of the sampling board 111, so that it is detected as a sampling signal by the photomultiplier 112. That is, referring to part (e) of FIG. 4, at the respective n-th, (n+1)-th and (n+2)-th sampling times, sampling signals P1, P2 and P3 are extracted by slit 109 of sampling board 111 and detected by the photomultiplier 112.
The sampling signals P1, P2, P3, . . . are detected in a time-series mode and are combined, as a result of which one pulse waveform of the incident light beam IN having the repetitive frequency can be observed with a predetermined time resolution.
In the case of FIG. 3, the sampling board 111 with the slit 109 is provided outside of the streak tube 130. However, as shown in FIG. 5, the sampling board may be provided inside the streak tube. That is, the sampling board may be arranged between the micro-channel plate 132 and the phosphor screen 134. In the streak tube 150 shown in FIG. 5, a sampling electrode 151 is provided therein and held at the same potential as the phosphor screen 134. The streak tube having the sampling electrode inside it is generally referred to as "a sampling streak tube."
With respect to the sampling streak tube, the sampling waveform can be extracted before the electron beam reaches the phosphor screen 134. Therefore, unlike the apparatus shown in FIG. 3 in which the electron beam is applied to phosphor screen 134 throughout the sweep operation, in an optical waveform observing apparatus with the sampling streak tube, the necessary electron beam is applied only to the central portion of phosphor screen 134, and only a part of the luminance distribution is applied to the photomultiplier 112. As a result, the sampling waveforms detected are less affected by the background noises of the phosphor screen.
In the streak tube 130 shown in FIG. 3, or in sampling streak tube 150 shown in FIG. 5, the accelerating electrode 135 is a circular mesh accelerating electrode and, therefore, the electron beam passing through the accelerating electrode 135 includes background noises. The mesh opening percentage of the accelerating electrode 135 is generally 60%. As shown in part (a) of FIG. 6, about 40% of the incident light beam which, not being subjected to photoelectric conversion by the photocathode 131, has passed therethrough, is scattered by the mesh material, to form scattered light beams DF. The scattered light beams DF thus formed are applied to photocathode 131 from behind and cause the photocathode to emit photoelectrons BG that represent the background noises.
Part (b) of FIG. 6 shows the density distribution of the electrons that reach the sampling electrode 151 in the streak tube 150. As is apparent from part (b) of FIG. 6, in the case where the accelerating electrode 135 is a mesh electrode, the electron density distribution .rho..sub.o and a scattered light beam photoelectron density distribution .rho..sub.e. The latter density distribution .rho..sub.e represents the background noises and lowers the sampling waveform observation accuracy. As seen from the density distribution .rho..sub.e, the photo electrons BGT, due to the scattered light beams, are greatest in number at the central portion of the phosphor screen to which the incident light beam is applied, and fewer in number towards the periphery of the phosphor screen. The substantial number of scattering electrons around the incident light beam thus lowers the measurement accuracy. As a result, it is difficult to accurately measure a light pulse waveform with high time resolution, since background noises due to the scattered light beams DF are produced before and after the measured light pulse.
The background noises as represented by the density distribution .rho..sub.e are spread over a wide range. As a result, during a non-sweep standby period, while the incident light beam is applied to the photocathode 131, the noise electrons of density distribution .rho..sub.e will be sampled with high probability. Accordingly, those noise electrons should be eliminated. For this purpose, in the conventional optical waveform observing apparatus, the deflecting voltage V applied by the deflecting circuit 108 is made large in amplitude and high in through rate (voltage/time). This can lower the probability that, during the electron beam deflection, the photoelectrons attributable to the scattered light beams are sampled.
As the amplitude of deflecting voltage V is made large, the potential V.sub.m of the deflecting voltage V is large during the non-sweep stand by period. Therefore, even if the incident light beam is applied to the photocathode 131, the signal electrons attributable to the incident light beam are greatly deflected by the deflecting electrode 133 and reach a peripheral portion of the phosphor screen 134 that is remote from the central portion. As a result, even if the density distribution .rho..sub.e of the photo electrons due to the scattered light beams is spread over a wide range, the probability that these photoelectrons reach the central portion of phosphor screen 134 is small and, accordingly, their effect on the sampling waveform is eliminated.
With respect to the sampling streak tube 150 shown in FIG. 5, the potential V.sub.m of the deflecting voltage V is made large enough so that, during the non-sweep stand by period the electron beam reaches a portion of the sampling electrode 151 that is remote from the slit. Accordingly, in this case also, the sampling waveform is protected from being affected by the photoelectrons caused by the scattered light beans. In practice, the potentials V.sub.m and -V.sub.m of the deflecting voltage V are about +1 KV and -1 KV, respectively.
In the optical waveform observing apparatus in which the waveform of an incident light beam is observed by sampling, an aperture time .DELTA.t in which the sampling waveform is extracted in determined by a sweeping velocity V.sub.t at which the electron beam emitted from the photocathode 131 in response to the incident light beam is deflected to sweep across the slit of the sampling electrode, a diameter u of the electron beam, and a slit width w of sampling electrode 151. The aperture time .DELTA.t corresponds to the time resolution and is represented by the following equation (1): ##EQU1## The sweeping velocity V.sub.t can be represented by the following equation (2): EQU V.sub.t =S.times.T
where S is the deflection sensitivity (cm/V) of the sampling streak tube 150, and T is the through rate (V/sec) of the deflecting voltage.
As is apparent from equation (1) and (2), the aperture time .DELTA.t is decreased as the deflection sensitivity S and/or the deflecting voltage through rate T increase, with the result that the time resolution is improved.
However, the conventional optical waveform observing apparatus experiences difficulties when the distance between the pair of deflecting electrodes 133 is decreased for improvement of the deflection sensitivity S. Since the deflecting voltage V.sub.m has been made large in amplitude for the reasons described above, the greatly deflected electron beam will, during the non-sweep stand by period, strike against and be reflected by the deflecting electrodes and the reflected beam will travel to the central portion of the phosphor screen 134. Accordingly, in the conventional optical waveform observing apparatus, it is impossible to greatly reduce the distance between the pair of deflecting electrodes 133. That is, the possible improvement to the deflection sensitivity S is limited. As a result, it is impossible to obtain a time resolution on the order of several picoseconds and, further, it is also impossible to increase the repetitive frequency of the deflection.
The above-described difficulties are eliminated by an apparatus disclosed in Japanese patent application No. 163535/87 entitled "Optical Waveform Observing Apparatus". A corresponding U.S. patent application Ser. No. 291,893 was filed on 12/29/88. This apparatus disclosed in those applications is as shown in FIG. 7.
In the optical waveform observing apparatus of FIG. 7, a sampling streak tube 1 comprises a photocathode 131 to which an incident light beam is applied, an accelerating electrode 2 for accelerating the electron beam emitted from the photocathode, deflecting electrodes 133 for deflecting in a predetermined direction the electron beam passed through the accelerating electrode 2, a sampling electrode 4 for sampling the electron beam deflected by the deflecting electrodes 133, and phosphor screen 134 for detecting the electron beam sampled by a slit 5 of the sampling electrode 4. Accelerating electrode 2 is in the form of a plate having an opening 3. Blanking deflecting electrodes 21 are interposed between sampling electrode 4 and deflecting electrodes 133 so that during a flyback period of the deflecting electrodes 133, the electron beam may not be sampled. The incident light beam applied to the photocathode 131 of the sampling streak tube 1 is a light pulse having a repetitive frequency whose waveform is to be observed. The light pulse is outputted by a pulsed light source 10 and is applied through a beam splitter 102, optical delay means 11, and an input optical system 12 to the photocathode 131.
In similar fashion to the beam splitter in the conventional optical waveform observing apparatus, the beam splitter 102 splits the incident light beam outputted by the pulsed light source 10 into two parts, one of which is applied to a photodiode 105, where it is converted into a deflecting trigger signal TR. Also, in similar fashion to the conventional apparatus, the delayed deflecting trigger signal TR is applied to a time sweep circuit 107, where it is gradually delayed so that the sampling operation is carried out once every sampling time as was described above. The delayed deflecting trigger signal TR is applied to a deflecting circuit 13 and a blanking deflecting circuit 22.
Photodiode 105 is used to form the electrical trigger signal TR by photo-electric conversion. However, a drive signal for pulsed light source 10 may be employed as the electrical trigger signal TR by operating a switch 14.
In the optical waveform observing apparatus constructed as illustrated in FIG. 7, The plate-shaped accelerating electrode 2 with opening 3 is employed and, therefore, the photoelectrons due to the scattered light beams are blocked by the solid (not opened) portion of the accelerating electrode. As a result, the probability that the photoelectrons are mixed with the signal electrons, thus providing background noises, is greatly reduced. Therefore, when the amplitude of the deflecting voltage applied across the deflecting electrodes 133 is reduced, the background noises caused by the incident light beam applied during the non-sweep stand by period are quite small. Accordingly, the probability that the background noises are sampled by the slit 5 of the sampling electrode 4 is reduced. Thus, the deflecting voltage applied to deflecting electrodes 133 can be made small in amplitude and the distance between deflecting electrodes 133 can be decreased, as a result of which the deflection sensitivity is improved and the deflecting repetitive frequency can be increased.
With respect to the optical waveform observing apparatus shown in FIG. 7, as a result of a employing plate-shaped accelerating electrode 2 with opening 3, the probability that the background noises caused by the photo electrons due to the scattered light beams being sampled during the non-sweep stand by peiod when the deflecting voltage V is at V.sub.m, is reduced to about 10.sup.-6 of the probability that the signal electron peak value is sampled. As a result, the deflection repetitive frequency can be made as high as about 4 MHz and, accordingly, the waveform of an incident light beam high in repetitive frequency can be observed.
In the case where the incident light beam repetitive frequency is much higher that the deflection repetitive frequency, e.g., on the order of 4 GHz, the frequency of application of the incident light beam to photocathode 131 of the sampling streak tube 1 during the non-sweep stand by period is quite high, and the background noises produced are substantially in proportion to that frequency.
Parts (a), (b) and (c) of FIG. 8 are time charts corresponding to the operation of the apparatus illustrated in FIG. 7 when the incident light beam repetitive frequency is much higher than the deflection repetitive frequency, such as described above. Specifically, parts (a), (b) and (c) of FIG. 8 are time charts showing the incident light beam IN, the deflecting voltage V, and the sampling waveform, respectively.
When, as shown in parts (a) and (b) of FIG. 8, the repetitive period T.sub.1 of the deflecting voltage V is much longer than the period T.sub.0 of the incident light beam IN, the deflection of the incident light beam IN is carried out once every T.sub.1 /T.sub.0 times. As a result, the sampling waveforms P1 and P2 are extracted with timing as shown in part (c) of FIG. 8. During the period T.sub.2 in which the deflection, i.e., the sweep, is not carried out, the deflecting voltage v is held at V.sub.m ' and the incident light beam IN is applied to the photocathode 131 of the sampling streak tube 1 about T.sub.1 /T.sub.0 times. The resultant background noises are extracted as sampling waveforms N as shown in part (c) of FIG. 8. That is, the sampling waveforms N are observed T.sub.1 /T.sub.0 times during the non-sweep stand by period. The magnitude of the background noise due to sampling waveforms N is much smaller than that of the signal electron sampling waveforms P1 and P2, being on the order of 10.sup.-6 of the latter. However, the integration of the former is (T.sub.1 /T.sub.0).times.10.sup.-6, which is so large that is cannot be disregarded with respect to the magnitude of the signal electron sampling waveforms P1 and P2.
A numerical example to illustrate this problem associated with background noise is described next. In the case where the repetitive frequency of the incident light beam IN is about 4 GHz and the repetitive frequency of the deflecting voltage is 4 MHz, the waveform of the incident light beam is deflected once every 1,000 applications of the incident light beam IN and sampled. However, the incident light beam is applied to the photocathode 131 of the sampling streak tube 999 times during the non-sweep stand by period during which no deflection is carried out. In this case, the integration of the background noise due to the sampling waveforms is on the order of 10.sup.-3 (=999.times.10.sup.-6) and it cannot be disregarded when high accuracy observations are to be performed.
In the case where the incident light beam is a continuous wave light beam, the background noises will greatly affect the signal electrons. If, in such a case, the aperture time (time resolution) is 10 psec (picoseconds) and the deflection repetitive frequency is 4 MHz (the repetitive period being 250 nanoseconds), then the integration of the background noise sampling waveforms N is about 0.025=(250.times.10.sup.-9 /10.times.10.sup.-12).times.10.sup.-6 of the signal electron peak value. This value cannot be disregarded when performing high accuracy observations.
In addition to the background noises resulting from the photo electrons caused by scattered light beams, the dark current of the photocathode 131 of sampling streak tube 1 and that of the photocathode of the photomultiplier 112 may adversely affect the accuracy of observation.